The present document defines a framework for use within ETSI NFV ISG to coordinate and promote public demonstrations of Proofs of Concept (PoC) illustrating key aspects of NFV.
The objective for the PoCs is to build commercial awareness and confidence and encourage development of an open ecosystem by integrating components from different players.
This framework outlines:
The present document provides an overview of NFV acceleration techniques and suggests a common architecture and abstraction layer, which allows deployment of various accelerators within the NFVI and facilitates interoperability between VNFs and accelerators. The present document also describes a set of use cases illustrating the usage of acceleration techniques in an NFV environment.
In accordance with the functional reference architecture of cloud computing that was defined in Recommendation ITU-T Y.3502, Recommendation ITU-T Q.3914 specifies the functional reference architecture of cloud computing according to Recommendation ITU-T Y.3500. This Recommendation provides a set of parameters that indicate the status and event of a cloud computing system, including resource layer, service layer and access layer.
The present document describes a study of how today's Cloud/Data Centre techniques can be adapted to achieve scalability, efficiency, and reliability in NFV environments. These techniques are designed for managing shared processing state with low-latency and high-availability requirements. They are shown to be application-independent that can be applied generally, rather than have each VNF use its own idiosyncratic method for meeting these goals. Although an individual VNF could manage its own scale and replication, the techniques described here require a single coherent manager, such as an orchestrator, to manage the scale and capacity of many disparate VNFs. Today's IT/Cloud Data Centres exhibit very high availability levels by limiting the amount of unique state in a single element and creating a virtual network function from a number of small replicated components whose functional capacity can be scaled in and out by adjusting the running number of components. Reliability and availability for these type of VNFs is provided by a number of small replicated components. When an individual component fails, little state is lost and the overall VNF experiences minimal change in functional capacity. Capacity failures can be recovered by instantiating additional components. The present document considers a variety of use cases, involving differing levels of shared state and different reliability requirements; each case is explored for application-independent ways to manage state, react to failures, and respond to increased load. The intent of the present document is to demonstrate the feasibility of these techniques for achieving high availability for VNFs and provide guidance on Best Practices for scale out system architectures for the management of reliability. As such, the architectures described in the present document are strictly illustrative in nature.
Accordingly, the scope of the present document is stated as follows:
Provide an overview of how such architectures are currently deployed in Cloud/Data Centres.
Describe various categories of state and how scaling state can be managed.
Describe scale-out techniques for instantiating new VNFs in a single location where failures have occurred or unexpected traffic surges have been experienced. Scale-out may be done over multiple servers within a location or in a server in the same rack or cluster within any given location. Scaling out over servers in multiple locations can be investigated in follow-up studies.
Develop guidelines for monitoring state such that suitable requirements for controlling elements (e.g. orchestrator) can be formalized in follow-up studies.
The present document reviews virtualisation technologies and studies their impact on the NFV architectural framework and specifications. It also provides an analysis of the pros and cons of these technologies.
The present document identifies the most common design patterns for using SDN in an NFV architectural framework. It also identifies potential recommendations to be fulfilled by the entities that perform the integration.
ETSI ISG NFV has defined an NFV architectural framework operating on the basis of the principle of separating network functions from the hardware they run on by using virtual hardware abstraction. The major components in this framework are (From ETSI GS NFV 002):
Network Functions Virtualisation Infrastructure (NFVI): subsystem which encompasses Compute, Network and Storage resources, i.e. the totality of all hardware and software components that build up the environment in which VNFs are deployed.
Management and Orchestration (MANO): subsystem which includes the Network Functions Virtualisation Orchestrator (NFVO), the Virtualised Infrastructure Manager (VIM) and Virtual Network Function Manager (VNFM).
Virtual Network Functions (VNFs): deployed in the NFVI.
The present document provides an overview of SDN in relation to this ETSI NFV architectural framework as well as a summary of current industry work including a comparison of network controllers and PoCs including NFV and SDN.
The present document develops a set of normative interoperability requirements for the Network Function Virtualisation (NFV) hardware ecosystem and telecommunications physical environment to support NFV deployment. It builds on the work originated in ETSI GS NFV 003.
The present document focusses on the development of requirements to enable interoperability of equipment in the telecommunications environment to support NFV deployment. The following areas are examined:
The present document focusses on the characterization of Virtualised Network Functions (VNF) as part of their configuration and deployment in "the Cloud". Such VNFs are assumed to be implemented using generic cloud computing techniques beyond virtualization [i.1]. For example, the VNFs can be built with re-usable components as opposed to a unique - and potentially proprietary - block of functions.
Cloud native VNFs are expected to function efficiently on any network Cloud, private, hybrid, or public. The VNF developer is therefore expected to carefully engineer VNFs such that they can operate independently in the desired Cloud environment. Cloud environment can be implemented based on hypervisor/VM or container technology. This is an indication of the "readiness" of VNFs to perform as expected in the Cloud. The objective of the present document is to develop the characterization of the "Cloud Readiness" of VNFs.
From an operator perspective, it is essential to have a complete description of cloud native readiness of VNFs; this description will help operators in their VNF selection process. To do this, it is essential that a set of non-functional parametric characterizations be developed that appropriately describe the cloud native nature of VNFs. Non-functional parameters describe the environmental behaviour of VNFs residing in the Cloud. They do not describe the actual working functions of the VNF; rather they describe how the VNF can reside independently in the Cloud without constant operator involvement.
The present document considers not only the "pure" cloud native VNF implementations (e.g. no internal resiliency or state) but also some transition implementations to cloud native such as the VNFs with internal resiliency.
Non-functional characteristics of a cloud native VNF are described through a VNF Product Characteristic Descriptor (VNFPCD) created by the VNF provider. Usage of the cloud native VNFPCD is as follows:
The cloud native VNFPCD is used by an operator to decide on what VNF product to deploy to fulfil a particular functionality, when the decision is based on non-functional parameters.
The VNFPCD can be used in a VNF market place for a standardized description of the VNF products non- functional characteristics and as such can be checked/searched for automatically.
The intent of the present document is to identify a minimum set of non-functional parameters by which VNFs are characterized as cloud native. The non-functional parameters are classified according to the specific environmental behaviour of the VNF.
Each behaviour then provides a list of specific non-functional parameters along with specific requirements such that the cloud native nature of the VNF can be satisfactorily established.
The present document specifies requirements for a set of abstract interfaces enabling a VNF to leverage acceleration services from the infrastructure, regardless of their implementation. The present document also provides an acceleration architectural model to support its deployment model.
The goals of the present document are:
to identify common design patterns that enable an executable VNFC to leverage, at runtime, accelerators to meet their performance objectives;
to describe how a VNF Provider might leverage those accelerators in an implementation independent way; and
to define methods in which all aspects of the VNF (VNFC, VNFD, etc.) could be made independent from accelerator implementations.
VNF providers have to mitigate two goals:
VNFs might have constraints to perform their function within certain power consumption boundaries, CPU core count, PCI express slot usage and with good price/performance ratio; and
VNFs should accommodate most if not all deployment possibilities.
The present document specifies performance benchmarking metrics for virtual switching, with the goal that the metrics will adequately quantify performance gains achieved through virtual switch acceleration conforming to the associated requirements specified herein. The acceleration-related requirements will be applicable to common virtual switching functions across usage models such as packet delivery into VNFs, network overlay and tunnel termination, stateful Network Address Translators (NAT), service chaining, load balancing and, in general, match-action based policies/flows applied to traffic going to/from the VMs. The present document will also provide deployment scenarios with applicability to multiple vendor implementations and recommendations for follow-on proof of concept activities.
The present document specifies functional requirements for both the Virtualised Infrastructure Manager (VIM) and the NFV Infrastructure (NFVI), for NFV acceleration from an infrastructure management perspective. This includes the controlling and management of acceleration resources, e.g. allocation, release and discovery of acceleration resources. The present document also identifies the corresponding impacts on VIM related specifications regarding functional requirements ETSI GS NFV-IFA 010 and reference points (ETSI GS NFV-IFA 005 and ETSI GS NFV-IFA 006).
The present document specifies the interfaces supported over the Or-Vi reference point of the NFV-MANO architectural framework ETSI GS NFV 002 as well as the information elements exchanged over those interfaces.
